**** BEGIN LOGGING AT Wed Aug 16 03:00:01 2017 Aug 16 15:31:43 pmezydlo: report? Aug 16 15:32:17 sorry, now I'm writing Aug 16 15:42:15 m_w: report is ready Aug 16 15:44:58 got it Aug 16 15:50:58 meeting in 10 minutes Aug 16 15:55:31 Also just writing my report up now Aug 16 15:56:23 maybe we should make the reports due on monday so they actually get done on tuesday like they are supposed to be Aug 16 15:58:44 m_w: some of us in my area had a get-together on sunday. christopher freidt came and brought the tiny beagle thing you made :-) Aug 16 15:58:50 <_av500_> aloo Aug 16 15:58:58 hi Aug 16 15:58:59 hello Aug 16 15:59:23 tlwoerner: cool! Aug 16 15:59:30 I am famous Aug 16 15:59:44 m_w: we're all looking forward to your lofive thing (offtopic) Aug 16 16:01:16 m_w: can confirm, held the little sucker in my hand Aug 16 16:01:17 super cool Aug 16 16:01:34 good evening everyone Aug 16 16:01:54 <_av500_> m_w: where is mine? Aug 16 16:01:57 <_av500_> and what does it do? Aug 16 16:02:06 yes, the point of the get-together was to meet with ee, it was great! i can't believe i never thought to take a group photo!! :-( Aug 16 16:02:17 Hello Aug 16 16:02:22 Hi everybody Aug 16 16:02:34 _av500_: it does nothing and everything all at once Aug 16 16:02:43 tlwoerner: ah, it's ok, we'll meet up when I get back and we'll get one in Aug 16 16:02:51 "make me one with everything" Aug 16 16:03:02 ee: +1 Aug 16 16:03:18 hi all :) Aug 16 16:03:20 I owe you a bottle(s) of something and I'll have to deliver it anyways ;) Aug 16 16:03:30 hi all Aug 16 16:04:16 hello :) Aug 16 16:04:32 me0w Aug 16 16:04:34 <_av500_> 1) welcome Aug 16 16:04:38 <_av500_> who's here? Aug 16 16:04:45 i'm here Aug 16 16:04:49 me Aug 16 16:04:49 i'm Aug 16 16:04:51 here Aug 16 16:04:56 I am here Aug 16 16:05:06 me too Aug 16 16:05:15 6! Aug 16 16:05:18 <_av500_> gee Aug 16 16:05:33 * _av500_ does a happy dance Aug 16 16:05:34 <_av500_> so Aug 16 16:05:38 <_av500_> final week is soon Aug 16 16:05:42 <_av500_> as in next week Aug 16 16:06:12 geez deadlines approaching fast Aug 16 16:06:29 <_av500_> yes Aug 16 16:06:34 <_av500_> I count 6 reports as well Aug 16 16:07:21 <_av500_> thanks :) Aug 16 16:07:25 <_av500_> so Aug 16 16:07:34 <_av500_> anybody with any burning issues? Aug 16 16:08:09 nope Aug 16 16:08:12 all good here Aug 16 16:08:16 me too Aug 16 16:08:22 Things are looking good over here. Lot's more free time now so no worries Aug 16 16:08:24 (yet) Aug 16 16:08:29 famous last words hahaha Aug 16 16:08:33 looking good Aug 16 16:08:56 it's fine Aug 16 16:09:01 <_av500_> good Aug 16 16:09:28 <_av500_> so without further ado, lets do the status reports Aug 16 16:09:36 <_av500_> who's this week? Aug 16 16:09:48 <_av500_> who was last? Aug 16 16:09:57 I am this week Aug 16 16:09:59 me afair with indu Aug 16 16:10:02 were last week Aug 16 16:10:21 I and ee this week Aug 16 16:10:31 i am up for next week then Aug 16 16:11:01 <_av500_> ravikp7 and ee ok Aug 16 16:11:03 Want to lead ravi? Aug 16 16:11:09 sure Aug 16 16:11:21 thanks :) Aug 16 16:11:33 no problem, starting.. Aug 16 16:12:45 So, my project is heading towards the final goal, that is a GUI flashing app for the BeagleBone Aug 16 16:13:02 Code here: https://github.com/ravikp7/BeagleBoot Aug 16 16:13:35 This week I integrated all etcher flashing modules in the app Aug 16 16:14:02 The app currently works on Linux and OSX Aug 16 16:14:27 Morning ! Aug 16 16:15:03 We can boot BeagleBone into USB mass storage mode and flash the emmc with the app now Aug 16 16:16:14 For booting into usb mass storage mode, using the bootloader server that I developed Aug 16 16:18:16 One of the main challanges in the app was to get sudo prompt working in the app to elevate only the scripts that require elevation. As we shouldn't ideally elevate the whole electron based app as it's a browser underneath Aug 16 16:19:34 ravikp7: can you provide some travis-ci pre-built image releases? Aug 16 16:19:35 For this to be working, I fork a child process from the main app process that then executes a script with elevation Aug 16 16:20:08 ravikp7: very nice on getting the sudo working. Aug 16 16:20:12 jkridner|pd: sure, after meeting Aug 16 16:20:51 So, we get two processes, the main app process with no sudo access, and other process with sudo access Aug 16 16:21:25 An IPC server is run in the main process and an IPC client in the other elevated process Aug 16 16:21:38 <_av500_> nice Aug 16 16:22:34 +1 Aug 16 16:22:55 that's pretty hardcore, nice work ravikp7 :) Aug 16 16:23:01 The main process having IPC server directs the other elevated process (IPC client) to run requested scripts along with neccessary arguments with sudo access Aug 16 16:23:11 how do you prevent the sudo able process from being tricked into doing anything ? Aug 16 16:24:29 ds2: the client emits message to server with socket in order to get the scripts info Aug 16 16:25:31 <_av500_> ds2: also, rot13! Aug 16 16:25:42 without socket info, the server process can't send messages to the client Aug 16 16:26:02 socket? Unix domain? INET domain? Aug 16 16:26:20 rot13? for what? Aug 16 16:26:37 unix Aug 16 16:26:43 if rot13 is good, rot26 must be better :D Aug 16 16:26:53 <_av500_> ds2: I use tripple rot13 Aug 16 16:27:15 ravikp7: so if I find it.... Aug 16 16:27:41 so the client first requets the server with server's id, then only it accepts messages Aug 16 16:28:19 I guess then, no other process can then excecute scripts with the elevated process Aug 16 16:29:17 The two scripts that needed sudo are usb mass storage script, and the image write script Aug 16 16:30:29 For Windows support, we first need to get the bootloader server working on Windows Aug 16 16:32:01 The problem there is that the spl device has two usb configs, and Windows set the default config which we don't require and setting other usb config is not possible with libusb on Windows Aug 16 16:32:57 So, we can try changing the configs, and if we get default as required on Windows then it can work Aug 16 16:34:16 ravikp7: are you still using the older bootloader? Aug 16 16:34:37 Coming to app, for Windows support I need to add some code for Windows as things get a bit different for Windows for elevation process (like checking sudo access and passing env variables) than unix Aug 16 16:35:08 m_w: yes, still no success with mainline Aug 16 16:35:28 jkridner: right? Aug 16 16:36:03 ravikp7: I've been playing with mainline, but haven't tried to edit the default USB configuration yet. Aug 16 16:36:19 ravikp7: I think there could be a work-around by issuing CONTROL_TRANSFERS directly. Aug 16 16:37:00 So, for now things are working good on Linux and OSX, the app works quite good there for flashing Aug 16 16:37:37 <_av500_> nice Aug 16 16:37:53 <_av500_> too bad for the niche OS :) Aug 16 16:38:02 ^ heh Aug 16 16:38:30 <_av500_> ravikp7: thanks Aug 16 16:38:40 I will be completing the app GUI, the settings and about page this week, along with windows support (for elevation) and some bug fixes Aug 16 16:39:00 finally then packaging apps for supported platforms Aug 16 16:39:19 <_av500_> ok Aug 16 16:39:30 <_av500_> thanks again Aug 16 16:39:33 I request all to give it a try.. it needs testing :) Aug 16 16:39:54 <_av500_> if jkridner sends me the BBpink Aug 16 16:40:07 funny enough you mention that my BBB is giving me all kinds of boot shenanigans so I'll give it a shot Aug 16 16:40:18 Thanks all for attention and appreciation Aug 16 16:40:43 I suppose it's my turn now Aug 16 16:41:22 <_av500_> yes Aug 16 16:41:43 Bit of a long week as I've been traveling and searching for an apartment, but I managed to complete the I2C implementation for the Go version of BeagleLibs. Aug 16 16:41:56 I whined about Go during my last weekly report but I take all of it back Aug 16 16:42:42 Compared to Rust, go is much more like C which means you dodge all the verbosity that comes along with Rust's safety guarantees. While this means the code is a lot cleaner and easier to write, it means that I'll have to do more testing to ensure it works as it should. Aug 16 16:43:27 I also took a different approach to the implementation. For Rust, everything was done through ioctl syscalls as opposed to using the i2cdev file for read/write operations Aug 16 16:44:03 This has it's performance benefits, but when combined with Rust's additional verbosity it made for code that's hard to read and a pain to write and reason about. Aug 16 16:44:26 I'm thinking of moving the Rust code to a similar approach as Go as part of this week's work. Aug 16 16:44:49 Other than that the next week will be busy. On the todo list is completing the SPI and UART implementation Aug 16 16:45:04 alongside a huge effort that will have to be made in the documentation and testing area Aug 16 16:45:50 I hope to have full examples for the hardware I was sent as part of the hardware shipment, for example the OLED display (I2C) and the fingerprint scanner (UART) Aug 16 16:46:17 as well as more basic things like circuit diagrams for the simpler stuff so new people can get into it quickly Aug 16 16:46:53 speaking of hardware I received my x15 this week (and a nice customs receipt) so I'm excited to get working with that. Aug 16 16:47:17 do you think you'll be able to look at adding the ability for the user to more easily pick which pin they want to use in the code and make that more intuitive? Aug 16 16:47:25 Overall, things are progressing and like I anticipated they're moving along much faster due to having experience in rust implementation Aug 16 16:47:47 tlwoerner: good that you brought that up, it slipped my mind after we spoke about it during our meet up Aug 16 16:48:24 Yes, that's something that hugely reduces the cognitive load on new users (as well me lol) so it's something I'll definitely implement Aug 16 16:48:51 great! Aug 16 16:48:52 I'm weighing how to approach it, but I have an idea of how I'll implement it Aug 16 16:49:31 for everyone who wasn't at the meeting, the basic idea is to implement an alias for the PWM/GPIO pins Aug 16 16:49:54 Currently when you're initializing a PWM you've got to supply a PWM number + a PWM chip number Aug 16 16:50:21 When you're configuring an output to be a PWM, you use the "header" name, i.e. P9.22 Aug 16 16:50:56 this is a barrier to usability, so just aliasing the PWM + PWM chip numbers to something friendlier should help Aug 16 16:51:07 <_av500_> keep in mind there is not much time left Aug 16 16:51:16 <_av500_> its important to wrap up next week Aug 16 16:51:23 <_av500_> that goes for everybody Aug 16 16:51:42 I'll just drop an example of what I meant: https://github.com/ekmecic/libbeaglebone/blob/master/examples/smooth_led.rs#L11 Aug 16 16:51:47 but yes, that's all for me Aug 16 16:51:49 thank you all :) Aug 16 16:51:54 <_av500_> thanks Aug 16 16:52:00 <_av500_> thanks everybody Aug 16 16:52:02 everything needs to be pushed to git, by next week, right? Aug 16 16:52:05 <_av500_> yes Aug 16 16:52:09 <_av500_> push it! Aug 16 16:52:44 dumb question, but is additional content (i.e. docs, examples) that's pushed after the deadline not up for consideration? Aug 16 16:53:45 <_av500_> well, its until consideration as long as the mentor has not graded it Aug 16 16:53:49 <_av500_> under* Aug 16 16:54:03 <_av500_> talk to your mentor Aug 16 16:54:30 Got it. I ask because I know the docs/examples will be revised all the time Aug 16 16:54:45 ee: we will coordinate so I don't grade too soon Aug 16 16:55:11 tlwoerner: awesome, sounds good Aug 16 16:56:27 is that all for the meeting? Aug 16 16:57:06 it's been an hou Aug 16 16:57:07 hour Aug 16 16:57:50 jkridner: which site did you use for the x15 shipments? Aug 16 17:23:17 thetransformerr[: still here? Aug 16 17:23:22 Hi, still someone here? Aug 16 17:25:14 Damn, couldnt make it earlier... Aug 16 18:34:28 ee: I've forgotten where I put the pre-orders in now. :( Aug 16 18:50:55 jkridner: I havn't done any travis-ci pre-built image release before, if you can provide some links, that would be helpful Aug 16 18:51:12 jkridner: like this https://github.com/AppImage/AppImageKit/wiki/Docker-Hub-Travis-CI-Workflow ? Aug 16 18:52:17 * jkridner looks at http://hoverbaum.gitlab.io/2016/08/03/Building-Electron-based-applications-using-CIs/ Aug 16 18:53:12 +1 Aug 16 18:54:08 Even Etcher seems to be using Travis-CI: https://github.com/resin-io/etcher Aug 16 23:43:05 hi wormo:, I just sent you a mail Aug 16 23:44:00 I fell asleeep, by the time, the meeting was to over, hence I missed your ping Aug 16 23:45:09 I can even remember that I was going to ping you, anyways, I have sent you a summary that you may like to review **** ENDING LOGGING AT Thu Aug 17 03:00:00 2017